Measurement and Shimming of the Improved CERN Synchro-Cyclotron Magnet

Abstract

All magnetic measurements of the CERN SC2 magnet were done with Hall-plates.The principal element of the measuring gear was a 3.5m long bar with 100 Hall-plates mounted on a computer-controlled turn-table. The Hall-voltages were scanned, converted into field values and stored on tape.For special regions other Hall-plate arrangements were used. The field maps were evaluated in terms of mean radius, frequency, n-value, etc. versus energy. The optimum field for capture near the centre, acceleration and transmission through the extraction system was calculated.Shimming was required near the centre and in the channel-regenerator region. The same programs as for the original design were used. The programs used will be briefly described. Shim CalculationThe reconstruction of the CERN synchrocyclotron was undertaken to get a considerable increase of the proton beam intensity. The principal changes were a new RF system giving higher Dee voltage and repetition rate and a new hooded arc ion source and central geometry to get efficient capture into a well-defined beam. The axial introduction of the ion source required a 20cm Top cone Washer 2010 10 20 diameter hole to be drilled through the poles.Simultaneously the magnet focusing near the source needed improving. These contradictory requests could only be fulfilled by an intricate shimming. Also the extraction system, that now consists of a regenerator, an electromagnetic septum and three iron channel sections, required shimming to minimize undesired perturbance of the orbits. The shimming was followed by magnetic measurements and reshimming in several cycles.Capture calculations had given an initial "ideal" field shape and "ideal" frequency curve. The field shape and pole shape should be unchanged outside 25cm radius. To calculate the shimming around the central hole a program called FUNCFLD was used 1 )2). The field in the gap is represented by a number of current loops. Another set of current loops together with a permeability function represent the field in the iron. The desired field in the midplane and co-ordinates of a set of known points on the pole surface (in this case outside 25cm radius and in the hole) can be specified. The program attempts to match scalar and vector potentials for gap and iron on a pole surface and give the desired field.At first every attempt failed; the reason was soon found to be that the large hole in the centre contradicted Maxwell's equations for the desired field. Every effort was then made to decrease the hole diameter, particularly near the pole surface (see Fig. 1